Answers are grounded in this paper and cite their sources.
As online payment systems advance, the total losses via online banking in the United Kingdom have increased because fraudulent techniques have also progressed and used advanced technology. Using traditional fraud detection models with only raw transaction data cannot cope with the emerging new and innovative schemes to deceive financial institutions. To increase the detection accuracy of machine learning and deep learning models, we propose a new feature-engineering framework that can produce the most effective feature set for any algorithm by embedding both methods of feature engineering and feature selection into a new framework. In the previous framework, we suggested using a selected dataset that includes new engineered features without any irrelevant features to a target label for any fraud detection algorithms. However, through investigating many related studies in financial fraud detection, we recognised that we selected many different types of machine learning or deep learning for building a fraud detection model. In the experiment, we proved the effectiveness of adopting our new framework by using real-life banking transactional data provided by a private European bank and evaluating the performance of the designed fraud detection models appropriately. Machine Learning and Deep Learning models perform at their best when working with the created feature set of the new framework.
No takes yet. Share an insight, caveat, or question.
Ouazzane et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: